Imelt problem in cadence simulation.

[iVenky]

While simulating the netlist didn't show any error regarding imelt. After doing parasitic extraction and including the "av_extracted" file, I am getting imelt error.


What is the reason behind this?

thanks in advance.

 

[erikl]

imelt: too large a current (imminent to melting something), either for a device or a routing wire. Occasionally occurring due to a device breakthrough event (on its part due to loading/unloading parasitic caps/inductivities), or due to (un)loading currents into/from large (parasitic) caps.

 

[dick_freebird]

imelt is just one of the "asserts" which can be reported by a
model instance. It may be realistic, arbitrary, or default.

Changing parasitics changes things like overshoots, and
the assert doesn't care whether your "melt" condition lasts
for an attosecond or a minute. Models which are not well
fitted, well beyond the application limits, can produce very
unreasonable results when a numerical interation steps out
of those bounds (as they will, in the course of searching
on either side of the solution).

It's worth peering at some of the reported timepoints*devices
just to see what the deal really is. Even though usually it's
just bogus whining.